Completely Random BSOD that I can't pin down.

Associate
Joined
18 Oct 2002
Posts
1,717
Location
Hemel
The last two weeks I've been getting the odd BSOD followed by POSTing failure direclty afterwards which I can't for the life of me pin point where the problem is.

It started about two weeks ago when I installed a USB Laser Printer onto my machine. Within an hour of that I got my first BSOD ever on this machine. I immediatly removed the printer and uninstalled the drivers hoping this was the cause. However the blue screens have stayed with me.

They aren't daily and follow no pattern. They can happen while gaming, web browsing or sometimes even while the PC is just idling. Sometimes it's just the one, sometimes it'ls multiple back to back.

My components are:
  • AMD Phenom II X6 Six Core 1090T Black Edition 3.20GHz
  • Corsair Hydro H50-1 High-Performance Liquid CPU Cooler
  • Asus M4A89TD PRO AMD 890FX (Socket AM3) DDR3 Motherboard
  • Corsair XMS3 8GB (4x2GB) DDR3 PC3-16000C9 2000MHz Dual-Channel Kit
  • Asus ATI Radeon HD 6970 DirectCU II 2048MB GDDR5
  • Cooler Master Silent Pro Modular 700W Power Supply
  • 2 x OCZ Vertex 2E 60GB 2.5" SATA-II Solid State Hard Drive in a Raid0
  • 2 x 500gb Seagate Barracuda in a Raid1 Mirror
  • 1 x 500gb Seagate Barracuda Stand Alone
  • Creative X-Fi

All of my components are at stock and have never been overclocked.
All BIOS settings are as default straight out the box which has worked perfectly since August 2010 when these parts were purchased.

The only hardware change recently aside from the printer was to switch from a 5850 to a 6970 at the beginning of April. As it was ATI to ATI I didn't reinstall Windows but I did uninstall ATI software and run DriverCleaner before reinstalling drivers for the new card. It ran fine for a month before the first BSOD.

So far I've tried the following to track it down:
  • Upped the RAM Volts to 1.65v which they're rated for rather than the 1.5v the BIOS gives it by default and checked the the timings are correct.
  • I've left memtest86 running for over 12 hours and found no problems with my RAM
  • I left Prime95 running for over 12 hours and no BSOD occured.
  • I left 3D Mark 2011 running on a loop for 6 hours without a BSOD
  • Run IntelBurnTest Multiple times without a BSOD
  • Run FurMark BurnIn without BSOD
  • Multiple AV Scans to check for nasties lurking
  • Monitoring Temps with CoreTemp and MSI Afterburner (all fine)
  • Bought a new PSU just incase it was the PSU failing - which didn't solve it

I can't replicate the problem through stress tests and I've updated every driver I can find! It's really starting to frustrate me.

It'll be fine for days and days then out of the blue for no reason it'll blue screen. After the blue screen I'll have to switch it right off for a minute of so before it'll post again. At which point the ASUS Bios will say "Overclock Failed" even though there is no overclock....unless it's doing it itself!!

An example of some of the BSOD's...

Code:
The problem seems to be caused by the following file: ahcix64s.sys

PAGE_FAULT_IN_NONPAGED_AREA

*** STOP: 0x00000050 (0xfffff683807de418, 0x0000000000000000, 0xfffff80002e619e7, 
0x0000000000000005)

*** ahcix64s.sys - Address 0xfffff88000dd77f3 base at 0xfffff88000db0000 DateStamp 
0x4adf0f64

Code:
The problem seems to be caused by the following file: ntoskrnl.exe

MEMORY_MANAGEMENT

*** STOP: 0x0000001a (0x0000000000041790, 0xfffffa80024753c0, 0x000000000000ffff, 
0x0000000000000000)

*** ntoskrnl.exe - Address 0xfffff80002e88700 base at 0xfffff80002e18000 DateStamp 
0x4d9fdd34

Code:
The problem seems to be caused by the following file: atikmdag.sys

*** STOP: 0xa0000001 (0x0000000000000005, 0x0000000000000000, 0x0000000000000000, 
0x0000000000000000)

*** atikmdag.sys - Address 0xfffff880058a8db1 base at 0xfffff8800585c000 DateStamp 
0x4dae3c99

Code:
The problem seems to be caused by the following file: ntoskrnl.exe

PAGE_FAULT_IN_NONPAGED_AREA

*** STOP: 0x00000050 (0xfffff68380792848, 0x0000000000000000, 0xfffff80002dc21f5, 
0x0000000000000005)

*** ntoskrnl.exe - Address 0xfffff80002cbb700 base at 0xfffff80002c4b000 DateStamp 
0x4d9fdd34
 
Last edited:
Last edited:
Boooo, I was hoping to avoid formatting :(

I've reseated everything except the CPU. When it refuses to POST it gets stuck on the RAM but no amount of memtest seems to find a problem with any of the sticks.
 
Oh forgot to add that to my list of things I've tried. Yes, I updated the BIOS last weekend. No change.

It just blue screened again a moment ago, all I did was open Spotify and Steam :(
All open windows stopped responding, but the mouse still worked, then the whole thing froze and blue screened.

Code:
The problem seems to be caused by the following file: ntoskrnl.exe

MEMORY_MANAGEMENT

*** STOP: 0x0000001a (0x0000000000041790, 0xfffffa8000b65260, 0x000000000000ffff, 
0x0000000000000000)

*** ntoskrnl.exe - Address 0xfffff80002e8d700 base at 0xfffff80002e1d000 DateStamp 
0x4d9fdd34
 
I'm running 4 DIMMs bt it was stable for over 6 months before the blue screens started :(

I've taken two sticks out for now to see if it improves while I start getting ready to do a format/reinstall
 
Last edited:
I did a search and it looks like quite a few people have encountered the PAGE_FAULT_IN_NONPAGED_AREA problem.

This guy ended up RMA'ing his Asus mobo and replacing it with a Gigabyte one instead:http://forums.overclockers.co.uk/showthread.php?t=18257714&highlight=PAGE_FAULT_IN_NONPAGED_AREA

That will be me :p

But yeah, OP, when it doesn't POST - turn it off, press the CPU heatsink lightly but firmly, and power it up. Bet you it POSTs. Mine always would, it got to a point where it would blackscreen while gaming, and I'd press the heatsink and voila, it'd come back to life as normal. Not that I recommend it while it's on of course :p

If it has issues POSTing it isn't to do with the windows install.

Oh yeah, tried Windows' inbuilt driver verifier too?
 
I'm going to confidently point the finger of blame at the RAM

Had exactly the same problem..turned out to be both sticks dead.RMA'd and all was well after
 
I'm going to confidently point the finger of blame at the RAM

Had exactly the same problem..turned out to be both sticks dead.RMA'd and all was well after

Wouldn't bad RAM show up on memtest over night?

I can't seem to force it to do it nor can I find anything wrong.
Memtest is clear after 12 hours and IntelBurnTest runs on every setting I throw at it without problems.

Its been fine the rest of the day. It has a little tantrum this morning but now it's fine. Annoying! It's not fixed, it'll probably do it again soon.
 
Wouldn't bad RAM show up on memtest over night?

I can't seem to force it to do it nor can I find anything wrong.
Memtest is clear after 12 hours and IntelBurnTest runs on every setting I throw at it without problems.

Its been fine the rest of the day. It has a little tantrum this morning but now it's fine. Annoying! It's not fixed, it'll probably do it again soon.

Sometimes it can pass memtest and still be failed, if the failing blocks are in the reserved section I believe.

Mine was the same, let us know what happens.
 
Well sadly I can comfortably say it's flakey hardware somewhere in my rig.

Last night I put all four sticks back in and started to perform a format/reinstall.
No problems throughout the install and I thought things were looking good.

Before bed I left it transferring WoW from my backup onto one of the SSDs.
When I came back to it this morning I find that it has blue screened over night :(

There are no drivers installed to cause problems and no software. This install is completely vanilla.

I can only assume it's the Memory based on the bluescreen messages.

Code:
The problem seems to be caused by the following file: ntoskrnl.exe
 
IRQL_NOT_LESS_OR_EQUAL
 
Technical Information:
 
*** STOP: 0x0000000a (0x0000000000000028, 0x0000000000000002, 0x0000000000000000, 
0xfffff800028bc090)
 
*** ntoskrnl.exe - Address 0xfffff80002879f00 base at 0xfffff80002808000 DateStamp 
0x4a5bc600

Code:
The problem seems to be caused by the following file: ntoskrnl.exe
 
PAGE_FAULT_IN_NONPAGED_AREA
 
Technical Information:
 
*** STOP: 0x00000050 (0xfffff683807ef080, 0x0000000000000000, 0xfffff80002986e77, 
0x0000000000000005)
 
*** ntoskrnl.exe - Address 0xfffff80002881f00 base at 0xfffff80002810000 DateStamp 
0x4a5bc600

Out of the four sticks I have NO idea which ones are the bad ones.
I'm currently running just two and seeing how it goes.
Previous Memtest runs didn't show any problems.

What confuses me is how I can go from a completely stable rig to this flakey machine through changing just a GFX card and installing a USB printer.

Obviously it looks like I'm going to have to replace something, but I'm reluctant to start buying anything until I know for certain that what I'm replacing is the faulty component.

Could 4x2gb of DDR3 in this motherboard coupled with the added power drain of a 6970 instead of the old 5850 be making the system unstable?
 
something may have popped. Try with each set of ram for a while. Best using matching stick though, but I would say a memory stick or the motherboard memory controller, or some incompatiblity between them, altohugh unlikely since the problem developped recently.
 
The trouble I'm having is prooving it.

I can't force the bluescreens to happen nor can I diagnose the faulty sticks, if they are indeed faulty. Memtest comes back clean every time and stress tests cope fine.

It's just a case of waiting for something the might not happen...very time consuming and completely random.
Annoying :P
 
It's been fine all day again today :(

I've tested both pairs on their own with memtest for a few hours and both passed fine.

I've also been playing WoW for that last 3 hours without any problems.

Sooo annoying!
 
Reading through this, you say you get a message that say "OC failed" even when not OC.. This is normally when then CPU fan stops working or the RPM is lower then the tolerance set in the BIOS system monitor. Check the the fan is 100% working at all times. Move the fan header to another power source and turn off CPU_FAN monitoring then perform a stress test. It could be the fan stops and periodical overheating may cause problems with the Memory controller, which of course is integrated in the CPU.

I would also certainly be thinking of reinstalling windows with a fresh set of chipset drivers (if not done so already).

The only other thing you could try is purchase a cheap second hand AM3 processor. I mean really the cheapest you can find just to eliminate the controller and therefore the CPU.
 
Reading through this, you say you get a message that say "OC failed" even when not OC.. This is normally when then CPU fan stops working or the RPM is lower then the tolerance set in the BIOS system monitor. Check the the fan is 100% working at all times. Move the fan header to another power source and turn off CPU_FAN monitoring then perform a stress test. It could be the fan stops and periodical overheating may cause problems with the Memory controller, which of course is integrated in the CPU.

I would also certainly be thinking of reinstalling windows with a fresh set of chipset drivers (if not done so already).

The only other thing you could try is purchase a cheap second hand AM3 processor. I mean really the cheapest you can find just to eliminate the controller and therefore the CPU.

I'll have to keep an eye on that. The HSF is a H50 so not a standard HSF combo but it's worth keeping an eye on just to rule it out.

I did a clean install in Saturday night and it blue screened while I was sleeping so I think that rules out bad chipset drivers :(
 
Well at least it does rule out the software :)

Sorry I missed the fact it's a H50.. However, it could be failing pump or fan.

Can you not stick the reference HSF on and go back to stock settings.. just to rule out the H50?
 
It's odd that memtest isn't thowing anything up.
I'd put good money on it being the creative card... they're always causing me issues.

Try a fresh install without the creative ever being used. See if you still get them.
 
Back
Top Bottom